The Rel/NFkB familiy of transcription factors is widely expressed and influences the expression of numerous genes, including many involved in immune and inflammatory responses. Understanding the regulation of this family is therefore an important step toward influencing these respones. We have focused on the role of NFkB in the differentiation of normal peripheral blood monocytes into cells possessing many of the phenotypic and functional properties of dendritic cells (DC). DCs are the most efficient of the body's antigen-presenting cells and hence are critical for proper immune system function. They are of special interest currently because of their possible therapeutic utility. To facilitate the study of signaling pathways involved in DC differentiation, we developed a serum-free culture system in which monocytes differentiate rapidly (48-96 hours) in response to bacterial lipopolysaccharide (LPS), tumor necrosis factor, or calcium ionophore. In spite of significant differences in the signaling pathways utilized by these agents, they all activated members of the NFkB family, including RelB. Furthermore, agents which prevented NFkB activation also blocked DC differentiation, suggesting that NFkB is required for the differentiation process. Using adenovirus as a model, we tested whether DC differentiation can be induced by infectious agents other than gram-negative bacteria. We found that infection of monocytes with replication-defective adenovirus triggered very rapid differentiation to cells with immuno- phenotypical, morphological, secretory, and functional properties of DCs. As in LPS-induced differentiation, this process involved, and appears to require, NFkB. The vigorous response of monocytes to replication-defective adenovirus offers a cautionary statement regarding the use of such vectors in gene therapy trials where the purpose is to install transgenes into human tissues while provoking minimal inflammatory responses. The strong pro-inflammatory response induced by the viral vectors suggests that the in vivo response of monocytes (and presumably other cell types) to infection with adeno-virus could have potentially negative consequences in human gene therapy settings.